Quick Answer: Solving How to Clean and Calibrate the Concept2 RowErg Flywheel & PM5 Sensor (2026)
The PM5 self-calibrates every stroke — you cannot manually calibrate it. Erratic drag factor means mechanical fouling: dust-packed flywheel housing, a glazed or misgapped sensor, or a stretched chain. Clean the screen, verify the sensor sits 1–2 mm from the flywheel magnets, confirm drag factor 100–135, and re-check every 250 hours.
The single most common misconception about the Concept2 RowErg is that the PM5 monitor requires periodic calibration in the way a torque wrench or a load cell does. It does not. The PM5 is a dead-reckoning computational device: it measures the angular velocity of the flywheel during the drive and, critically, the rate of deceleration of that flywheel during the recovery. From that deceleration curve it derives the moment of inertia opposing the rower and expresses it as the drag factor. Because every single stroke recalculates this value from first principles, there is no stored calibration constant to drift, corrupt, or lose. What does drift is the mechanical and electrical environment the monitor is measuring — and that is what you actually maintain. For detailed technical setup and system integration, reference our analysis on replace concept2 rowerg return shock cords (2026).
Root Cause Engineering & Mechanical Principles
The RowErg’s resistance system is an air-braked centrifugal fan. Resistance scales with the cube of flywheel angular velocity, which is why the machine feels self-correcting across athlete power outputs. The damper lever does not change resistance directly — it rotates a sliding plate over the intake vents on the flywheel housing, altering the mass of air admitted per revolution. More air mass means more work per stroke and a steeper deceleration curve during recovery. The PM5 samples that curve and reports drag factor, a unitless number that on a clean Model D or E typically ranges from roughly 65 (damper 1) to 220 (damper 10).
The sensing side is deliberately primitive and therefore extremely reliable. A ring of permanent magnets is bonded to the inner face of the flywheel. A passive pickup — a sensor coil terminating in a two-conductor cable that runs up the monorail to the monitor arm — counts magnet transits. Pulse interval gives angular velocity. There is no optical element, no encoder disc, and no analog gain stage to drift. Failures are therefore binary and mechanical: the gap is wrong, the cable is broken, the connector is dirty, or ferrous debris has bridged the pickup face.
Dust ingress is the dominant long-term failure mode. The flywheel housing acts as a centrifugal separator, flinging airborne particulate onto the inside of the screen and into the vane roots. A heavily loaded screen restricts intake, so a machine that read drag factor 130 at damper 5 when new may read 155 at the same lever position after three years in a carpeted garage. The athlete perceives this as “the erg got harder” — it did, but the wattage math remains correct because the PM5 is measuring actual physics, not assuming it. Scores stay valid; only the damper-to-drag mapping has shifted. For competitive context, the World Rowing Federation and most sanctioned indoor regattas govern verification by drag factor, not damper position, precisely because the lever number is meaningless across machines.
Two secondary systems corrupt the measurement indirectly. A stretched or kinked chain introduces non-uniform torque delivery, producing a sawtooth force curve and inconsistent stroke detection at the catch. A failing bungee shock cord — the recovery return elastic — allows handle slack, which the PM5 can misread as an early drive onset, inflating stroke rate and depressing split. Neither is a sensor fault, but both present to the athlete as “bad data.”
| Observed Symptom | Mechanical / Electrical Failure Mode | Diagnostic Check | Difficulty |
|---|---|---|---|
| Drag factor climbed 15–30 points at same damper setting | Flywheel screen and vane roots packed with dust; intake area reduced | Visual inspection through housing vents; compare drag at damper 10 against factory 200–220 reference | Easy — 20 min |
| Drag factor reads 0 or blanks mid-piece | Pickup cable chafed at monorail entry, or jack partially unseated at monitor arm | Continuity test across the 2.5 mm jack: expect <2 Ω conductor, >10 MΩ conductor-to-shield; flex cable while probing | Moderate — 45 min |
| Monitor wakes but shows no stroke until 5–6 hard pulls | Excessive sensor air gap (>3 mm) weakening induced pulse amplitude | Feeler gauge or 1.5 mm shim between pickup face and flywheel magnet ring; target 1–2 mm | Moderate — 30 min |
| Erratic stroke rate spikes to 60+ spm | Ferrous swarf bridging pickup face, generating spurious transits | Inspect pickup with bright light; wipe face with lint-free cloth and isopropyl | Easy — 15 min |
| Split times inconsistent stroke to stroke at steady effort | Stretched bungee (>10% elongation) causing handle slack and false catch detection | Measure handle return force; handle must retract fully and briskly with no sag at the chain guide | Moderate — 40 min |
| Grinding or ticking on recovery | Dry or spalled idler/sprocket bearing | Spin flywheel by hand with chain detached; listen for roughness, check for radial play >0.5 mm | Hard — 90 min |
| Drag factor artificially low on an otherwise clean machine | Obstruction over intake (towel, cover) reducing air mass — also the classic tampering vector | Remove all drapes; re-measure at damper 10 and compare to reference band | Trivial |
| Monitor loses power or resets under load | Battery contact corrosion; vibration-induced intermittent | DC voltage at terminals under load: healthy pair reads >2.9 V, replace below 2.6 V | Easy — 10 min |
Step-by-Step Diagnostic & Calibration Procedure
Work through this sequence in order. Each step establishes a precondition for the next, and skipping ahead will produce ambiguous results.
Step 1 — Establish a Baseline Drag Factor Reading
Before touching anything, capture the machine’s current state. From the PM5 main menu select More Options u2192 Display Drag Factor. Row ten firm strokes at approximately 22 spm with the damper at 10. Record the stabilized value. A healthy, clean Model D or E will land between 200 and 220. A reading above 230 indicates heavy intake fouling; below 185 suggests an obstruction, a non-standard flywheel, or an older Model B/C with the monitor type misconfigured. Repeat at damper 5 and damper 1 and log all three. These three points are your diagnostic signature — recheck them after every intervention below.
Step 2 — Confirm Monitor Machine-Type Configuration
A PM5 moved between machines retains its previous configuration, and a monitor set to SkiErg or BikeErg will compute nonsense drag factors on a RowErg. Navigate to More Options u2192 Utilities u2192 Machine Type and confirm it reads Indoor Rower. Then check Utilities u2192 Product ID for firmware revision. Keep firmware current via ErgData or the Concept2 Utility; several historical revisions corrected stroke-detection edge cases at very low drag factors. This is the closest thing to a software calibration the PM5 has.
Step 3 — Deep-Clean the Flywheel Housing
Unplug the sensor cable at the monitor arm before removing any shroud hardware so you are not working against a live, swinging lead. Remove the flywheel housing screws — on most Model D units these are #2 Phillips self-tapping screws into plastic bosses. On reassembly, torque them to approximately 0.8–1.0 Nu00b7m, or in practical terms: seat the screw until the shroud is flush and give it no more than a further eighth-turn. These bosses strip easily and there is no repair short of a heli-coil or adhesive.
With the shroud off, vacuum the loose dust first — do not blow it into the bearings. Then work compressed air at no more than 30 psi (2 bar) from the inside outward through the screen, holding the nozzle at least 150 mm from the magnet ring. Higher pressures can drive debris into the bearing seals and, on older machines, unseat magnet adhesive. Wipe the screen and the vane roots with a dry lint-free cloth. Avoid solvents on the flywheel itself; residue attracts the very dust you just removed.
Step 4 — Verify and Set the Sensor Air Gap
Locate the pickup on its bracket adjacent to the flywheel magnet ring. Target clearance is 1–2 mm. Below roughly 0.75 mm you risk physical strike as the flywheel finds its running position; above 3 mm the induced pulse falls below the monitor’s detection threshold at low rpm, which is exactly why a marginal machine “wakes up” only after several hard strokes. Set the gap with a 1.5 mm feeler gauge or a folded business card stack of known thickness, then rotate the flywheel a full revolution by hand and confirm the gap holds — magnet rings are not perfectly concentric and you are setting the minimum clearance, not the average. Snug the bracket fastener to roughly 1.5 Nu00b7m. Clean the pickup face with isopropyl alcohol on a lint-free swab and let it flash off fully before reassembly.
Step 5 — Electrical Continuity Verification
With the cable disconnected at both ends, set your multimeter to the 200 Ω range and probe tip-to-tip across the sensor lead. You are looking for a stable low-resistance path, under 2 Ω for the conductor itself. Then measure conductor-to-shield on the megohm range: anything under about 10 MΩ indicates insulation breakdown, typically where the cable passes the monorail entry and has been abraded by the seat carriage. Flex the cable along its full length while watching the display — an intermittent that only appears under flex is the signature failure, and it will present to the athlete as random mid-piece dropouts that are impossible to reproduce on a bench. Replace the sensor cable assembly rather than splicing; the shield integrity matters more than the conductor. For detailed technical setup and system integration, reference our analysis on clean & lubricate rowerg chains with 20w oil (2026).
Step 6 — Chain and Drive Train Service
Measure chain stretch by laying a steel rule along the chain under light tension: 20 links should span a nominal pitch distance, and elongation beyond 2% warrants replacement. Apply a light mineral-based chain oil — roughly 10 drops distributed along the chain, worked in by cycling the handle through ten full strokes, then wipe all surplus off with a shop towel. Excess oil is worse than none: it migrates onto the sprocket, collects grit, and forms a lapping compound. Service interval is every 250 hours of use or every 50 days, whichever comes first.
Step 7 — Bungee Tension Check
Pull the handle out to full extension and release it under control. It must retract fully and briskly to the chain guide with no sag and no bounce-back. If the handle stops short or drifts, the shock cord has lost elasticity — typically after 10% permanent elongation. Re-tension at the adjustment knot or replace the cord. Correct bungee tension is a stroke-detection prerequisite, not a comfort item.
Step 8 — Re-Baseline and Set Working Drag Factor
Reassemble, reconnect the sensor lead at the monitor arm, and repeat Step 1. Your damper-10 reading should now sit within the 200–220 band. Now set your actual training drag factor by lever position rather than by the number on the damper. The practical working range for most athletes is drag factor 100–135, which on a clean machine lands somewhere around damper 3–5. Heavyweight athletes and sprint-specific work trend toward 130–140; lightweight athletes and long steady-state sit comfortably at 100–115. Damper 10 is not “harder training,” it is a different, more strength-dominant stroke profile with a markedly longer recovery deceleration — and it is the fastest way to irritate a lumbar spine. Record your target drag factor, not your damper number, so the setting transfers to any machine anywhere.
Tools, Calibration Equipment & Component Replacement
The full service kit is modest. You need a #2 Phillips driver, a 3 mm and 4 mm hex key set, a feeler gauge set spanning 0.5–3 mm, a digital multimeter with continuity and megohm ranges, a vacuum with a brush attachment, compressed air regulated to 30 psi, lint-free cloths, 99% isopropyl alcohol, and a light mineral chain oil. A small torque screwdriver covering 0.5–2.5 Nu00b7m is a worthwhile addition if you service multiple machines — plastic boss stripping is the single most common self-inflicted damage during a first teardown.
Consumables and their realistic service lives: shock cord, 3–5 years or 2,000 hours; chain, 2,000–3,000 hours depending on lubrication discipline; sprocket, generally twice chain life; idler bearings, 5,000 hours or when audible roughness appears; monitor batteries, 12–18 months of regular use. The sensor pickup itself and the flywheel magnets are effectively lifetime components — if either has genuinely failed, the machine has usually suffered impact damage or water ingress.
Order genuine parts. The drag factor computation depends on the moment of inertia of the specific flywheel and the airflow geometry of the specific housing; aftermarket shrouds, non-standard screens, and third-party flywheel covers all shift the damper-to-drag mapping and in some cases invalidate the PM5’s internal model entirely. This is the one domain where parts commonality arguments do not apply.
Preventative Maintenance & Long-Term Reliability Protocols
Build the schedule around hours of use rather than calendar time, with calendar as a backstop. After every session: wipe the monorail with a dry cloth — perspiration salt is mildly corrosive to the aluminum and abrasive to the seat rollers. Weekly: wipe down the handle, seat, and frame; glance at the chain for dry spots. Every 50 hours or 50 days: oil the chain and check the bungee return. Every 250 hours: full flywheel housing clean, sensor gap verification, and a logged three-point drag factor baseline. Annually: inspect bearings, check all fasteners, and replace monitor batteries proactively.
Keep a maintenance log with the three-point drag signature from Step 1. This is the highest-value habit in the entire protocol. A single drag factor reading tells you nothing; a trend line tells you everything. If your damper-10 number has crept from 208 to 224 over eighteen months, you have a quantified intake fouling rate and you can schedule the clean before performance data gets noisy. It also makes the machine dramatically easier to sell — a documented service history on a RowErg holds resale value better than almost any other piece of home gym equipment.
Environment matters more than usage volume. A machine in a carpeted basement or a chalk-heavy gym floor will foul three to five times faster than one on sealed concrete in a clean room. If you cannot control the environment, halve the cleaning interval. Never store the erg with a cover draped over the flywheel housing — it traps moisture against the magnets and, if left in place during a session, drops the drag factor well below what the athlete believes they are rowing at.
Frequently Asked Questions
Can I manually calibrate the PM5, and is there a hidden service menu?
No, and no meaningful one. The PM5 recalculates drag factor from flywheel deceleration on every single recovery phase, so there is no persistent calibration constant to adjust. The only configuration that affects measurement validity is the machine type setting under Utilities, which must read Indoor Rower. Everything else you can influence is mechanical. If your numbers look wrong, the answer is in the housing or the cable, never in a menu.
Why does my drag factor read 160 at damper 5 when the chart says it should be around 120?
Intake fouling, almost certainly. Dust packed into the flywheel screen and vane roots reduces effective air admission per revolution, steepening the deceleration curve and raising drag factor at every lever position. Published damper-to-drag charts assume a factory-clean machine. Perform the Step 3 deep clean and re-baseline. If the number stays high after a thorough clean, check for a partial obstruction over the intake vents or confirm you are not on an older model with different housing geometry. Note that your past scores remain perfectly valid — the PM5 measured the real work you did. For detailed technical setup and system integration, reference our analysis on pair bluetooth heart rate straps to pm5 monitors (2026).
The monitor shows no data until I take several hard strokes. Is the sensor failing?
That specific signature — no reading at low rpm, correct reading once the flywheel is spinning fast — is the classic excessive air gap symptom. The induced pulse amplitude scales with flywheel speed, so a gap beyond about 3 mm falls below the detection threshold at the low angular velocities present in the first few strokes. Reset the gap to 1–2 mm per Step 4 before suspecting the pickup or cable. If the problem persists at a verified gap, move to the Step 5 continuity test and flex the cable at the monorail entry while probing; intermittent shield breakdown there is the next most likely cause.
